The data describes 2.5 million Expresso clients.

The objective of this hackathon is to develop a predictive model that determines the likelihood for a customer to churn - to stop purchasing airtime and data from Expresso.

The train file is large. We recommend making small splits on the train for local testing and not running one model on the whole train set.

Files available for download

• Train.csv - contains information about 2 million customers. There is a column called CHURN that indicates if a client churned or did not churn. This is the target. You must estimate the likelihood that these clients churned. You will use this file to train your model.
• Test.csv - is similar to train, but without the Churn column. You will use this file to test your model on.
• SampleSubmission.csv - is an example of what your submission should look like. The order of the rows does not matter but the name of the user_id must be correct.